Find the customers who stopped buying

What's inside

• Orders. Paste your order export: date, order number, customer email, total and any refund. A Check column catches duplicates, missing fields and customers not on your list.
• Customers. One row per customer. Orders, total spent, average order, first and last order, days since the last one and their usual gap between orders are all calculated for you.
• Segments. VIP, Active, New, At risk, One-time buyer, Lapsed, and VIP versions of at risk and lapsed, so you start with the customers worth the most.
• “Past usual reorder time”. A personal flag. With the default 1.5× setting, someone who normally buys every 30 days is flagged from day 45, even before they count as at risk.
• Win-back results. Add the date you contacted someone. If they order on or after that date, they're marked as won back, and the Dashboard shows your win-back rate.
• Room for 5,000 orders and 2,000 customers. Check columns flag duplicate orders and emails, dates the sheet can't read, and customers missing from your list.
• Dashboard. Repeat customer rate, share of revenue from repeat customers, average order value, orders per customer, and each segment's size and value.

Getting your data in from Shopify

1. Orders → Export as CSV. Copy the Created at, Name (the order number), Email, Total and Refunded Amount columns into the Orders tab. Shopify's dates come with a time and time zone (2026-09-01 14:22:05 -0400). Paste them as they are, and the sheet reads the date part.
2. Orders with several products come out as several rows, and the extra rows have a blank Total. Delete those rows so each order appears once. The Check column flags any order number that appears twice.
3. Customers → Export. Copy the Email column into the Customers tab, and Accepts Email Marketing into “OK to email?”. Shopify's customer export has no last-order date, which is why the orders are needed.

Other platforms work the same way. You need one row per order with a date, a customer email and a total.

Choosing your windows

When a customer counts as “at risk” depends on how often people normally buy from you. A coffee subscription and a sofa store are very different. The Settings tab shows your customers' average gap between orders, and Klaviyo's advice is to set the window a little longer than your average buying cycle. The defaults are 120 days for at risk and 240 for lapsed. Treat them as a starting point, not a rule.

Before you email anyone

• US: under CAN-SPAM you don't need prior consent, but every marketing email must, among other things, use honest From and subject lines, identify itself as an ad, include your postal address, and have a working unsubscribe that you honour within 10 business days. Penalties can be over $50,000 per email.
• UK: you generally need consent, or the “soft opt-in”: people whose details you got when they bought, or asked about buying, from you, receiving marketing for your own similar products, with an opt-out offered when you collected their email and in every message. See the ICO's guidance, which is being updated after the Data (Use and Access) Act 2025. EU countries have similar, but not identical, rules.
• The simplest rule: only email customers marked “Yes” in OK to email?.

Good to know

• “Repeat customers” here means customers with two or more orders in the file. Shopify's returning customer rate is measured over a date range, so the numbers won't match exactly.
• Average order value here uses order totals including shipping and tax, net of refunds. Shopify's AOV is gross sales minus discounts, without shipping and tax, so expect Shopify's figure to be lower.
• Want to know your profit per order as well? Use the e-commerce profit tracker.